An upcoming report from the National Academy of Sciences wraps up months of debate on a sticky issue for the bioinformatics community: What data and materials must authors of scientific papers provide upon publication, and in what manner should these materials be provided?
The report, “Sharing Publication-Related Data and Materials: Responsibilities of Authorship in the Life Sciences,” is the product of an initiative that the NAS launched in November 2001 to address growing uncertainty within the life science community regarding issues of data release and access in the post-genomic era.
The rise of high-throughput biological techniques — and access to the resulting data — has placed scientific publishers at the core of some controversial decisions in recent years, prompting NAS to nip the debate in the bud by identifying what, if any, “community standards” exist for making experimental results and research materials available upon publication. A panel of 10 representatives from academia, biotech, pharma, the funding agencies, and the scientific publishing community officially kicked off its discussion with a workshop last February [BioInform 03-04-02].
Nearly a year later, the committee is releasing what it considers to be some basic ground rules to guide authors, journals, and funding bodies. “Basically what we’ve come up with is some principles and some recommendations … just to remind everyone that this is how publication started — the quid pro quo thing — you publish your information, you acknowledge that you’ve done this, but you put that information out there,” said committee member Mary Waltham, former president and publisher for Nature, who currently runs her own consulting firm. “That system has really worked for a long time,” said panelist Sean Eddy, a bioinformaticist at Washington University. “I think that most people know that, but there’s been a temptation — especially with these big data sets — to backslide on that because you can’t get the datasets into the printed publication, so it lets people find loopholes.”
Arriving at a consensus wasn’t an easy process: “You are not dealing with a homogeneous world…so finding a common ground is actually quite difficult,” Eddy said. “This is one of those things where every thing you write down, you can think of an exception, and that’s what happened, and that’s why it took so long to get this report out the door."
At the heart of the report lies one concept that the committee was able to agree on. Termed UPSIDE (the uniform principle for sharing integral data and materials expeditiously), it is based on the idea “that the publication of scientific information is intended to move science forward.” In a nutshell, UPSIDE states that authors of scientific papers are obligated to release their publication-related data and materials so that others can replicate findings as well as build upon the published work. All members of the scientific community have equal responsibility for upholding these standards, according to UPSIDE.
The short-and-sweet acronym “struck us as a simple way of synthesizing what all this is about,” said Waltham. In addition, she noted, it may offer an easy way for members of the scientific community to align with the principles identified by the committee. “It would be wonderful if they all said, ‘We’re UPSIDE-compliant,’” she said.
A simple way for journals and funding agencies to pledge adherence to the principles outlined in the report may be crucial to enforcing them. According to the authors, scientific journals are currently doing a very poor job of communicating their current policies. In a survey of the 56 most frequently cited life science and medical journals, the report notes that only 41 percent offered a clearly stated policy on depositing data, 39 percent offered a policy regarding the sharing of materials, and only 2 percent had a policy on sharing software. There was no policy information at all for 45 percent of the journals. Most revealingly, a statement of consequences for non-compliance was only available for 2 percent of the surveyed journals.
Eddy noted that any compliance with the report’s recommendations would be purely voluntary — “There’s no enforceability in something like this,” he said — but added that one of the intended effects of the report is more transparency in journal policies. “The journal editors really didn’t have consistent policies. They’d like consistent policies, but they didn’t really know what the community wanted. ...So I think we’ll see more clarity in the journal policies as the result of this.”
The report provides 10 recommendations to navigate authors, journal editors, and funding agencies through the finer points of upholding the UPSIDE policy, as well as several hypothetical case studies highlighting potentially controversial scenarios and recommendations for how they might be addressed.
A pre-publication copy of the report is available at www.nap.edu/catalog/10613.html.